Anthrax vaccine powder formulations for nasal mucosal delivery · 5a. CONTRACT NUMBER 5b. GRANT NUMBER 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) Jiang /G Joshi, SB Peek, LJ Brandau,

Anthrax Vaccine Powder Formulations forNasal Mucosal Delivery

GE JIANG,1 SANGEETA B. JOSHI,2 LAURA J. PEEK,2 DUANE T. BRANDAU,2 JUAN HUANG,1

MATTHEW S. FERRITER,1 WENDY D. WOODLEY,1 BRANDI M. FORD,1 KEVIN D. MAR,1 JOHN A. MIKSZTA,1

C. ROBIN HWANG,1 ROBERT ULRICH,3 NOEL G. HARVEY,1 C. RUSSELL MIDDAUGH,2 VINCENT J. SULLIVAN1

1BD Technologies, 21 Davis Dr., RTP, North Carolina 27709

2Department of Pharmaceutical Chemistry, University of Kansas, 2095 Constant Ave., Lawrence, Kansas 66047

3US Army Medical Research Institute of Infectious Diseases, 1425 Porter St., Fort Detrick, Maryland 21702

Received 4 March 2005; revised 23 June 2005; accepted 4 August 2005

Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/jps.20484

ABSTRACT: Anthrax remains a serious threat worldwide as a bioterror agent. A second-generation anthrax vaccine currently under clinical evaluation consists of a recombinantProtective Antigen (rPA) of Bacillus anthracis. We have previously demonstrated thatcomplete protection against inhalational anthrax can be achieved in a rabbit model, byintranasal delivery of a powder rPA formulation. Here we describe the preformulationand formulation development of such powder formulations. The physical stability of rPAwas studied in solution as a function of pH and temperature using circular dichroism(CD), and UV-visible absorption and fluorescence spectroscopies. Extensive aggregationof rPA was observed at physiological temperatures. An empirical phase diagram,constructed using a combination of CD and fluorescence data, suggests that rPA is mostthermally stable within the pH range of 6–8. To identify potential stabilizers, a library ofGRAS excipients was screened using an aggregation sensitive turbidity assay, CD, andfluorescence. Based on these stability profiles, spray freeze-dried (SFD) formulationswere prepared at pH 7–8 using trehalose as stabilizer and a CpG-containingoligonucleotide adjuvant. SFD formulations displayed substantial improvement instorage stability over liquid formulations. In combination with noninvasive intranasaldelivery, such powder formulationsmay offer an attractive approach formass biodefenseimmunization. � 2005 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci

95:80–96, 2006

Keywords: anthrax; recombinant protective antigen; nasal; powder formulation;stability

INTRODUCTION

Anthrax, the zoonotic disease caused by theGram-positive spore-forming bacterium, Bacillusanthracis, can infect humans via cutaneous, gas-trointestinal, and pulmonary routes. The inhaledform is of particular concern considering its de-monstrated use as a bioweapon.1–4 Inhalational

anthrax is difficult to diagnose and effectivelytreat with antibiotics; therefore, prophylactic vac-cination is critical to elicit protection for indivi-duals at risk of exposure.5,6

B. anthracis produces a three component exo-toxin (anthrax toxin complex), which consists ofprotective antigen (PA, 83 kDa), edema factor (EF,89 kDa), and lethal factor (LF, 90 kDa).7–9 Whileall three proteins are nontoxic individually, thecombination ofPAandLF (designated lethal toxin)causes systemic shock and death associated withhyperoxidative burst and cytokine release frommacrophages.10,11 A combination of PA and EF,

80 JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 1, JANUARY 2006

Correspondence to: Vincent J. Sullivan (Telephone: 919-597-6173; Fax: 919-597-6402; E-mail: [email protected])

Journal of Pharmaceutical Sciences, Vol. 95, 80–96 (2006)� 2005 Wiley-Liss, Inc. and the American Pharmacists Association

Approved for public release. Distribution is unlimited.

Report Documentation Page Form ApprovedOMB No. 0704-0188

Public reporting burden for the collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering andmaintaining the data needed, and completing and reviewing the collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information,including suggestions for reducing this burden, to Washington Headquarters Services, Directorate for Information Operations and Reports, 1215 Jefferson Davis Highway, Suite 1204, ArlingtonVA 22202-4302. Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to a penalty for failing to comply with a collection of information if itdoes not display a currently valid OMB control number.

1. REPORT DATE 1 JAN 06

2. REPORT TYPE N/A

3. DATES COVERED

4. TITLE AND SUBTITLE Anthrax vaccine powder formulations for nasal mucosal delivery, Journalof Pharmaceutical Science 95:80 - 96

5a. CONTRACT NUMBER

5b. GRANT NUMBER

5c. PROGRAM ELEMENT NUMBER

6. AUTHOR(S) Jiang /G Joshi, SB Peek, LJ Brandau, DT Huang, J Ferriter, M Woodley,WD Ford, BM Mar, KD Mikszta, JA Huang, CR Ulrich, R Harvey, NGMiddaugh, CR Sullivan, VJ

5d. PROJECT NUMBER

5e. TASK NUMBER

5f. WORK UNIT NUMBER

7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) United States Army Medical Research Institute of Infectious Diseases,Fort Detrick, MD

8. PERFORMING ORGANIZATIONREPORT NUMBER

9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR/MONITOR’S ACRONYM(S)

11. SPONSOR/MONITOR’S REPORT NUMBER(S)

12. DISTRIBUTION/AVAILABILITY STATEMENT Approved for public release, distribution unlimited.

13. SUPPLEMENTARY NOTES The original document contains color images.

14. ABSTRACT Anthrax remains a serious threat worldwide as a bioterror agent. A second-generation anthrax vaccinecurrently under clinical evaluation consists of a recombinant Protective Antigen (rPA) of Bacillusanthracis. We have previously demonstrated that complete protection against inhalational anthrax can beachieved in a rabbit model, by intranasal delivery of a powder rPA formulation. Here we describe thepreformulation and formulation development of such powder formulations. The physical stability of rPAwas studied in solution as a function of pH and temperature using circular dichroism (CD), and UV-visibleabsorption and fluorescence spectroscopies. Extensive aggregation of rPA was observed at physiologicaltemperatures. An empirical phase diagram, constructed using a combination of CD and fluorescence data,suggests that rPA is most thermally stable within the pH range of 6-8. To identify potential stabilizers, alibrary of GRAS excipients was screened using an aggregation sensitive turbidity assay, CD, andfluorescence. Based on these stability profiles, spray freeze-dried (SFD) formulations were prepared at pH7-8 using trehalose as stabilizer and a CpG-containing oligonucleotide adjuvant. SFD formulationsdisplayed substantial improvement in storage stability over liquid formulations. In combination withnoninvasive intranasal delivery, such powder formulations may offer an attractive approach for massbiodefense immunization.

15. SUBJECT TERMS

16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF ABSTRACT

SAR

18. NUMBEROF PAGES

17

19a. NAME OFRESPONSIBLE PERSON

a. REPORT unclassified

b. ABSTRACT unclassified

c. THIS PAGE unclassified

known as edema toxin, induces an increase inintracellular cAMP and also elicits edema at thesite of infection.12,13 PA plays an essential role inanthrax pathogenesis, involving transport of LFand EF into the cytoplasm.4,14,15 PA is also theprimary immunogenic molecule that confers pro-tection against anthrax.16,17 The current licensedanthrax vaccines for human use are producedfrom the sterile culture supernatant fraction ofB. anthracis and are adsorbed on aluminumhydroxide for intramuscular (IM) injection18 intheUnitedStates andprecipitatedwithaluminumphosphate for subcutaneous (SC) injection in theUnited Kingdom.19 These vaccines contain vary-ing amounts of natural PA and also small amountsof LF and EF.20 A limited duration of protectionand a consequent need for frequent boosts tomaintain immunity have resulted in recent effortsto develop alternative vaccines such as those thatemploy recombinant PA (rPA) to improve theimmunogenicity, safety, and pharmaceutical con-sistency.2,21,22

A second generation anthrax vaccine consistingof a recombinantly produced PA antigen is cur-rently under clinical investigation where it isbeing administered by IM injection.23 Alternativeroutes other than IM could also be effective. Deli-very of vaccine to the respiratorymucosamay offeran attractive alternative to parenteral delivery forprotection against pulmonary anthrax infection.Intranasal (IN) immunization is efficient at elicit-ing mucosal immunity in the lungs,24–26 and hasbeen shown in animals to be an effective means ofimmunizing against pulmonary anthrax infec-tion.2 IN vaccination offers advantages, as it isnoninvasive route of delivery, with potential forself-administration. Nasal immunization withrPA could be further facilitated by the use of stablepowder formulations delivered via a unit dose,disposable device. Nasal powder delivery mayhave several advantages over conventional liquidformulations for parenteral routes including pro-tection from hydrolysis in bulk solution, potentialelimination of the need for refrigeration, increasedportability, and better suitability for mass immu-nization campaigns through an all-in-one prefilleddevice.

Previously, we reported that IN delivery of apowder formulation of rPA provides completeprotection against inhalational anthrax in a rabbitmodel.27 These results are summarized in Table 1.In vivo protective efficacy was reflected by animalsurvival after aerosol anthrax spore challenge.The IN powders elicited similar or slightly higher

protection compared to liquid formulations admi-nistered IMor IN (Tab. 1). In the present study, wedescribe the preformulationwork, including inves-tigation of the biophysical stability of rPA and thesubsequent formulation and characterization ofsuch IN rPA vaccine powders. The stability of rPAin solution as a function of pH and temperaturewas assessed and interpreted by an empiricalphase diagramapproach.28 In addition, stabilizerswere examined in order to identify a range ofoptimal formulation conditions. Stabilizers andprocessing on the intrinsic stability of rPA as wellas on its storage stability under accelerated condi-tions was also investigated.

MATERIALS AND METHODS

Materials

rPA was obtained from two different sources, theUS ArmyMedical Research Institute of InfectiousDiseases (USAMRIID, Fredrick, MD) and ListBiological Laboratories (Campbell, CA). rPA fromboth sources was expressed and purified fromBacillus anthracis and was equivalent based onthe criteria of size exclusion chromatography(SEC) and SDS–PAGE analysis. The CpG-con-taining oligonucleotide (#1826, TCCATGACGTT-CCTGACGTT) was purchased from Proligo, LLC(Boulder, CO). Chitosan (MW 50000–190000) waspurchased from Sigma (St. Louis, MO). CpG waschosen because it has been shown to be an effec-tive mucosal adjuvant in previous preclinicalstudies.29 All other reagents including buffersand salts were purchased from Sigma (St. Louis,MO) and were of analytical grade.

Table 1. Protective Effect from Intranasal DryPowder Formulations in Rabbits Receiving AnthraxInhalational Challenge (Each Rabbit Received 50 mgof rPA, n¼ 6 for Each Group)27

Formulation and Administration RouteAnimal

Survival (%)

Positive control (rPA/CpG liquid, IM) 83rPA/CpG liquid, IN 67rPA/CpG/trehalose SFD powder, IN 83rPA/CpG/trehalose SFD

powderþ chitosan, IN100

rPA/CpG/trehalose FD powder, IN 100rPA/CpG/trehalose FD

powderþ chitosan, IN100

Negative control (CpG) 0

ANTHRAX VACCINE POWDER 81

JOURNAL OF PHARMACEUTICAL SCIENCES, VOL. 95, NO. 1, JANUARY 2006


Experimental Methods

Physical Characterization of rPA Stability

rPA Sample Preparation. rPA solution was pre-pared at a concentration of 2.4 mM for UV ab-sorption and CD studies or 1.2 mM for fluorescencestudies in 50mMcitrate phosphate buffer contain-ing 0.1MNaCl at six different pH values from 3 to8 at one pH unit intervals.

UV Optical Density Measurements. The effect oftemperature on rPA aggregation was determin-ed by monitoring turbidity at 360 nm at eachpH value using an Agilent 8453 UV-Visiblespectrophotometer over a temperature range of10–908C at 2.58C intervals as described pre-viously.28 A 5-min equilibration time was employ-ed at each temperature interval to ensureequilibrium conditions.

Circular Dichroism (CD). CD studies were per-formed as described earlier30 using a Jasco J-810spectropolarimeter equipped with a 6-positionPeltier temperature controller tomeasure changesin the secondary structure of rPA as a function oftemperature. Each sample was examined in a0.1 cm pathlength cuvette sealed with a Teflonstopper, and the CD signal of rPA at 222 nm wasmonitored as a function of temperature from 108Cto 908C, collecting data every 0.58C. The CD signalwas converted to molar ellipticity using the JascoSpectra Manager software.

Intrinsic Tryptophan (Trp) Fluorescence. Intrinsicfluorescence measurements were performed asdescribed earlier30 to assess perturbations in thetertiary structure of rPA as a function of tempera-ture and pH. Fluorescence emission spectra of rPA(1.2 mM)were recorded from 10 to 858C using a PTIQuanta Master Spectrofluorometer equipped witha turreted, four-cell thermostatically control-led holder. An excitation wavelength of 295 nm(>95% Trp Emission) was employed and emissionspectra were collected from 305 to 440 nm. Exci-tationandemissionslitswere set at4nmanda1 cmpathlength quartz cuvette was used in all experi-ments. Spectra were collected at 2.58C intervalswith a 5-min equilibration time at each tempera-ture. Buffer baselines were subtracted from eachspectrumprior to data analysis usingFelixTM (PTI)software.Emissionpeakpositionswere obtainedby

a ‘‘center of spectral mass’’ method, which calcu-lates the zero moment of the entire spectral curvefrom 305 to 440 nm. Typically, the ‘‘center ofspectralmass’’ valuesdonot correspond toobservedpeak maxima (lmax), but do accurately reflect thetrend in actual peak position changes.

ANS Dye Binding Studies. The apolar dye, 8-Anilino-1-naphthalene sulfonate (ANS), oftenbinds to hydrophophic regions on proteins as theirconformation is altered. This usually results inenhanced fluorescence as well as a blue shift in theemission peak position. Thus, the association ofANS with rPA (1.2 mM) was monitored as a func-tion of temperature (10–858C) and pH (3–8) usinga PTI Quanta Master Spectrofluorometer. ANSwas added in a 20-fold molar excess over rPA. Anexcitation wavelength of 375 nm was used, andspectra were collected from 400 to 600 nm. Thetemperature ramping profile and instrument set-tings were the same as those used for the intrinsicTrp fluorescence experiments. Blanks of buffercontaining ANS were subtracted from each spec-trum prior to data analysis.

Constructing Empirical Phase Diagram

An empirical phase diagram was constructedas described previously28 but in this case usingCD molar ellipticity, intrinsic Trp fluorescenceemission peak center of mass, and ANS fluo-rescence intensity data sets. All calculationswere performed using the software Mathematica(Wolfram Research, Champaign, IL). The theoryand calculation process are described in more de-tail elsewhere.28 Briefly, experimental data setsare represented as n-dimensional vectors in atemperature/pH phase space, where n refers tothe number of variables (different types of data)included in the calculation. The data from eachtechnique at individual values of pH and tem-perature serve as the bases for the individualvector’s components. An n�n density matrixcombining all of the individual vectors is thenconstructed and n sets of eigenvalues and eigen-vectors of the density matrix are calculated. Thecomplete data set is subsequently truncated andre-expanded into three dimensions consisting ofeigenvectors corresponding to the three largesteigenvalues. The phase diagram is then gener-ated using vectors reconstructed in these threedimensions employing an arbitrary red/green/blue color scheme.

82 JIANG ET AL.



Screening for Stabilizers

An initial high-throughput screen for stabilizerswas performed utilizing the susceptibility of rPAto aggregation by monitoring the turbidity of thesolution with an optical density measurement at360 nm using a 96-well plate reader (BMG Lab-technologies, Offenburg, Germany). To identifypotential stabilizers, more than 30 ‘‘Generally Re-garded As Safe’’ (GRAS) excipients including buf-fers, sugars, amino acids, and surfactants werescreened for their ability to inhibit the aggrega-tion of rPA. In a 96-well plate, rPA (0.25 mg/mL)was added to wells containing each compoundfrom the GRAS library at selected concentrationsin citrate–phosphate buffer at pH 5. The sampleswere incubated at 378C, and the optical density ofthe solutions in each well was monitored at360 nm (OD360) for 70 min with readings obtainedat 5-min intervals. Controls of rPA solutionwithout excipient were examined simultaneously.The final OD360 observed at 70 min was used tocalculate the percent inhibition of rPA aggrega-tion produced by each compound.

Several effective inhibitors of rPA aggregationwere further studied to determine their effect onthe conformational stability of rPA upon thermaldenaturation, using the CD and intrinsic fluores-cence methods described previously. Titrationexperiments were performed with the selectedstabilizer by monitoring turbidity, as describedabove, to obtain the optimal concentration range ofthe stabilizer.

Formulation of Anthrax Dry Powder Vaccine

The anthrax dry powder was formulated by freezedrying (FD) or spray freeze drying (SFD) withrPA, trehalose, and an adjuvant comprising theCpG-containing oligonucleotide (#1826, TCCAT-GACGTTCCTGACGTT) (abbreviated herein asCpG). rPA was supplied at pH 8 in eitherammonium acetate solution (USAMRIID mate-rial) or in a freeze dried form (List Lab material)containing HEPES buffer salts. The rPA solutionwas mixed with CpG and trehalose solutions toobtain the final concentrations of 0.5 mg/mL rPA,0.5 mg/mL CpG, 99.0 mg/mL trehalose (with traceamount of salts from the rPA material). Part ofthis solution was stored at �708C for stabilitytesting and the remainder was aliquotted for FDand SFD. To produce SFD powder, the solutionwas first sprayed into liquid nitrogen using a BDAccuspray1 nozzle. The frozen particles obtain-ed were collected into lyophilization vials and

then transferred to a freeze drier (Dura-Stop/Dura-Dry, FTS Systems, Stonebridge, NY) pre-cooled to �408C. Finally, both FD and SFDsamples underwent a lyophilization cycle in whichthe primary drying was performed at�108C (shelftemperature) for 16 h, followed by a gradualramping of the shelf temperature to 208C forsecondary drying. The vials were stoppered undera nitrogen backfill. The FD cake was milled byball milling (Wig-L-Bug Grinding Mill, ReflexAnalytical Corp, Ridgewood, NJ) for 1 min at3000 rpm to generate the final powder. These rPA/CpG/trehalose powders can be filled into thecapsule of an IN delivery device previouslydescribed.27,31 Additionally, to determine if a bio-adhesive further enhances immunity, aliquots ofthe above powders were blended with a chitosanpowder prior to device filling.

Determination of rPA Stability duringthe SFD Process

The integrity of rPA throughout the SFD processwas determined by SDS–PAGE under both re-ducing and nonreducing conditions and by SEC-HPLC. Samples were collected at each step of theSFD process, that is, after mixing of rPA withCpG and trehalose, after the solution was sprayedfrom the nozzle, after the sprayed droplets werefrozen in liquid nitrogen, and after the finallyophilization. The final dried powder was alsoreconstituted in water for SEC-HPLC analysisfor comparison to freshly prepared rPA/CpG/trehalose solution.

Storage Stability Studies of Anthrax VaccineFormulations under Accelerated Conditions

The SFD anthrax vaccine powder was aliquot-ted into 2 mL serum vials (Wheaton, Millville, NJ)in a dry box (Terra Universal, Anaheim, CA)under nitrogen purge. Prior to filling with powdersamples, rubber stoppers and glass vials remain-ed overnight under nitrogen purge in a dry box.Each vial contained �25 mg of powder. The vialswere sealed with rubber stoppers and aluminumcrimp seals. The liquid formulation containing thesame composition as the SFD powder was alsoincluded in the stability study. After storage at258C or 408C, powder samples were reconstitutedin �0.25 mL water for analysis. Ten microliters ofeach sample were used for SDS–PAGE analysisand the remaining solution was filtered through a0.22 mm centrifugal filter (Millipore, Bedford, MA)prior to SEC-HPLC.




Size Exclusion Chromatography (SEC)

SEC was performed using a Waters HPLC instru-ment equipped with a binary pump (Waters 1525,Milford, MA), an autosampler (Waters 717 plus),and UV spectrophotometric detector (Waters 2487)interfaced to a Waters Breeze software package.Two size exclusion columns were used (Shodex1,JM Science, Inc., Grand Island, NY), a KW-803(cutoff 700000) column coupled to a KW-802.5(cutoff 150000) column to enhance resolution.Samples (75 mL) were analyzed by UV detection(l¼ 220 nm) employing a mobile phase consistingof 50 mM NaCl, 5 mM HEPES, 200 ppm sodiumazide, and a flow rate of 0.5 mL/min. After passingthrough the UV detector, the samples were addi-tionally analyzed using a multi-angle light scat-tering detector (Wyatt Technology Opitilab DSP,Santa Barbara, CA) for molecular weight mea-surement. The light scattering data were anal-yzed using Wyatt Astra software.

SDS–PAGE

SDS–PAGE was performed with a precastNuPAGE 4–20% TrisGly gel (Invitrogen, Carls-bad, CA). Protein samples and standards weremixed with 2� Tris-Glycine SDS sample buffer(Invitrogen) containing 2% SDS. For reducingconditions, dithiothreitol (1 M) and 2-Mercap-toethanol were added at 5% (v/v) each in thesamples. Samples were incubated for 8 min at958C, and electrophoresis was performed at a con-stant voltage of 125 V and 8 mA. The gel wasstained with Colloidal Blue (Invitrogen) and des-tained with distilled water.

RESULTS AND DISCUSSION

Physical Characterization of rPA Stabilityas a Function of pH and Temperature

The effect of pH and temperature on the physicalstability of rPA was analyzed using a number ofbiophysical approaches. First, aggregation of rPAwas both temperature- and pH-dependent as ob-served by monitoring the turbidity at 360 nm(Fig. 1A). Between pH 4 and 7, the temperatureat which aggregation began increased with pH.At pH 8, aggregation of rPA began at slightlylower temperature than at pH 7 with more sub-stantial reductions in aggregation onset tem-perature at lower pH. Additionally, the extent oftemperature-induced aggregation as indicated by

the maximum OD was lower for rPA at pH 7–8 than at pH 4–6. The subsequent decline of ODis indicative of extensive protein precipitation.At pH 7–8, turbidity onset occurred at highertemperatures suggesting that rPA is thermallymore stable in this pH range. At pH 3, aggrega-tion of rPA was only detected at temperaturesabove 808C, indicating that the aggregationmechanism may differ from that at pH 4–8.

The CD spectra of rPA display minima at about206–208 nmwith a shoulder at 216–217 nm at allpH values (Fig. 1B). This suggests that rPA pri-marily consists of amixture of secondary structuremoieties (e.g., a-helix and b-sheet). The calculatedsecondary structure from the CD spectra is com-parable to that of PA reported previously.32 At108C, the CD spectrum of rPA at pH 3 remaineddistinct from that at other pH values with a signi-ficantly lower molar ellipticity, indicating thatthe secondary structure of rPA in solution isaltered at very low pH. The secondary structureof rPAunderwent significant pHand temperature-dependent transitions, suggesting a loss of second-ary structure as evidenced by a decrease in thenegative ellipticity (Fig. 1C). The opposite is seen,however, at pH 3. Thermally induced ellipticitychanges occured at significantly lower tempera-tures at acidic pH values (e.g., pH 4) suggestingthat rPA is structurally more labile in an acidicenvironment. The different behavior seen at pH 3suggests that rPA undergoes a different type ofstructural alteration compared to that occurringat higher pH values (Fig. 1C). This is consistentwith the altered CD spectrum seen even at lowtemperature at this pH (Fig. 1B). No majorthermally induced changes were seen at pH 7–8,implying higher thermal stability under theseconditions; this is also consistent with the tem-perature/turbidity studies (Fig. 1A).

Intrinsic (Trp) fluorescence studies were per-formed to detect changes in the tertiary structureof rPA. The shift in fluorescence emission peakposition as a function of temperature at variouspH values is shown in Figure 1D. At all pH values,the peak position displays a red shift at elevatedtemperatures. This corresponds to an increase inexposure of the Trp residues to a more polarenvironment implying protein unfolding. Again,the onset temperature for red shift increased withincreasing pH. Thus, these experiments alsosuggest that the most stable environment for rPAis within the pH range of 7–8.

At pH 4 and 5, changes in the tertiary structureof rPA were detected in fluorescence spectra at

84 JIANG ET AL.



approximately 20–258C. In contrast, CD thermaltransitions began at higher temperatures, that is,338Cand 378C for pH4 and 5, respectively (Tab. 2).This suggests that a range of temperatures existin which rPA possesses a retention of pronounc-ed secondary structure and compactness, but apartial loosening of the native, tightly packedtertiary structure. Similar characteristics are alsoobserved for rPA at pH 6–8. This suggests theappearance of molten globule-like states33,34

between these temperatures. Such states are oftenprone to protein aggregation.35,36 The turbidityprofiles for rPA (Fig. 1A) support this idea since thetemperature at which aggregation begins for rPAat pH 4–8 falls within the temperature range

of molten globule-like states as determined byintrinsic fluorescence and CD.

The hydrophobic dye, ANS, was utilized as anextrinsic fluorescent probe to further investigatealterations in rPA tertiary structure. As a proteinbegins to unfold, the increased accessibility ofthe dye to the protein’s apolar interior oftenresults in an enhanced fluorescence signal. Insome cases, however, the negative charge on ANSis also thought to be involved in its interactionwithproteins. ANS is also believed to have a muchstronger affinity for protein ‘‘molten globulestates’’ than for native and severely structurallydisrupted conformational states.37–39 The changein ANS fluorescence intensity as a function of

Figure 1. Physical stability of rPA as a function of pH and temperature in citrate–phosphate buffer as determined by using optical density to detect insoluble aggregates(A); CD spectra of rPA at 108C (B); CD thermal melts monitored at 222 nm (C); intrinsicTrp fluorescence spectra center of mass (D); ANS dye binding fluorescence intensity (E).




Figure 1. (Continued )

86 JIANG ET AL.



temperature and pH is shown in Figure 1E.Conformational transitions appear to follow thesame patterns observed in the previous data. Forexample, at pH 3.0, rPA displayed significantapolar character even at 108C, signified by a highANS fluorescence intensity. This suggests thatthe protein is already structurally altered, per-haps to a molten globule state. At pH 4–8, theonset of increase in fluorescence intensity gener-ally shifted to a higher temperature as pHincreased. Also noteworthy is the decrease in theamplitude of fluorescence intensity as the pHincreased, suggesting less binding of ANS to rPAat increasing pH. These findings again demon-strate that the optimal pH range for rPA stabilityis 7–8.Moreover, theANSfluorescence results areconsistent with the presence of molten globulestates at intermediate temperatures (Tab. 2).

An Empirical Phase Diagram

A phase diagram was constructed using the re-sults of ANS fluorescence intensity, the intrinsictryptophan fluorescence center of mass and theCD intensity. Such a color map provides informa-tion regarding the state of rPA’s secondary andtertiary structures as a function of both pH andtemperature. This mathematical approach pro-vides a useful tool with which to interpret largequantities of data from different biophysicaltechniques for wide pH and temperature ranges.Each color represents a different physical state ofthe protein. Thus, a region of consistent colorcorresponds to an identifiable state while abruptcolor differences correspond to changes in proteinstates (i.e., phase boundaries). The nature of eachphase can be identified by simultaneously looking

at the original CD and fluorescence thermal meltspectra. We emphasize the empirical nature ofsuch diagrams (no equilibrium between states isimplied) and their primarily practical utility asrepresentations of protein behavior in solution.

In the phase diagram for rPA (Fig. 2), drasticcolor contrast clearly reflects changes in conforma-tional stability over a variety of pH and tempera-ture ranges. According to the previous discussion,the red-colored region in the lower, right-handcorner of the diagram can be identified as thephase ofmaximum stability (i.e., the ‘‘native’’ stateof the protein). A second phase (blue/purple) isobserved at pH 3 at temperatures below 458C andalso appears to encompass pH 4 & 5 at tempera-tures between 25–408C and 30–458C, respecti-vely. As discussed previously, this state possessessignificant molten globule character. At highertemperatures at pH 3, rPA rapidly enters anotherstate reflected by the dark brown color. This phaserepresents a severely structurally altered state ofthe protein. In the pH range of 4–7 at hightemperatures (>508C) and pH 8 (50–658C), thelight purple to light brown zone corresponds tothe structurally altered and aggregated states,

Table 2. Thermal Transition Onset Temperaturesof rPA Characterized by Optical Density (OD360nm),CD, Intrinsic Trp Fluorescence, and ANSDye-Binding Fluorescence

PH

Biophysical Approaches

OD360 CDTrp

FluorescenceANS Dye-Binding

Fluorescence

3 81.78C 16.48C 16.58C 10.08C4 30.48C 32.88C 22.58C 21.58C5 36.48C 36.58C 25.28C 30.18C6 43.78C 43.08C 37.58C 42.38C7 46.88C 48.08C 42.18C 45.08C8 47.98C 46.38C 40.08C 40.88C

Standard errors for all runs were 0� 0.58C. Figure 2. Phase diagram of rPA based on intrinsic,ANS dye-binding fluorescence, and CD results. Distinctphases are observed: (1) most stable phase [red-coloredregion in the lower right-hand corner]; (2) moltenglobule-like state [blue/purple area at pH 3, <458C,pH 4, 25–408C, and pH 5, 30–458C]; (3) severelystructurally altered phase [dark brown area at pH 3,>458C]; (4) structurally altered and aggregated state[light purple� light brown region at pH 4–7,>508C andpH 8, 50–658C]; (5) highly structurally disrupted form[green area at pH 8, >658C]. Blocks of continuous colorrepresent single phases, conditions underwhich the rawdata-derived vectors behave similarly.




which were identified previously in the UV turb-idity studies. At pH 8, yet another phase exists at>658C, which we speculate may correspond to aneven further structurally disrupted form. Thisphase diagram serves to identify the boundaries atwhich changes occur and can be used to selectconditions for the development of a high-through-put stabilizer screening assay and for formulationoptimization.

Screening for Stabilizers

As shown in Figure 1A, aggregation is a pro-minent pathway of physical degradation for rPA.Therefore, a turbidity-based high throughputscreening assay was developed to examine theability of various GRAS excipients to inhibit theaggregation of rPA. Based on the phase diagram,screening was performed at pH 5 and 378C. Theseconditions were chosen based on the intrinsicmarginal stability of rPA at this pH and tempera-ture. The conditions serve well the requirementsof accelerated stability testing since the proteinreadily aggregates at this pH and temperature. Atthe same time, they are sufficiently moderate tobe relevant to phenomena occurring under typicalpharmaceutical storage conditions. Figure 3 showsa representative plot of aggregation versus timefor rPA at 378C with and without the presence ofthree stabilizing excipients (malic acid, Tween-20,and sodium citrate). Several compounds exhibiteda significant (>50%) ability to suppress rPA

aggregation (Tab. 3). In contrast, pluronic F-68and dextran T-40 at certain concentrations in-duced rPA aggregation. The disaccharide treha-lose was found to be one of the most effectiveaggregation inhibitors. The extent of inhibi-tion of rPA aggregation was also concentration-dependent (Fig. 4). In this case, 5% or higherconcentrations of trehalose elicited �50% inhibi-tion of rPA aggregation.

After certain aggregation inhibitors were iden-tified, a number were further studied for theireffects on the conformational stability of rPAusingCD (Fig. 5A) and intrinsic Trp fluorescence(Fig. 5B). The data indicate that citrate, mannitol,and trehalose stabilize both the secondary andtertiary structure of rPA against thermal pertur-bation while Brij 35 has little effect. Trehaloseproduced the greatest stabilization, as demon-strated by a significant elevation in Tm of 108C inboth CD and intrinsic fluorescence studies.

Formulation of Anthrax Dry Powder Vaccine

Having identified certain conditions that lead toenhanced stability of rPA in the liquid state, weprepared and characterized dry powder formula-tions based on these findings. These powders havebeen previously shown to protect rabbits frominhalational anthrax challenge27 (Tab. 1). Thepowders were produced by either FD or SFDprocesses. The lyophilized cakes did not displayany evident collapse or obvious shrinkage. The

Figure 3. Aggregation of rPA in the presence and absence of excipients including0.15 M malic acid, 0.01% Tween 20, and 0.2% sodium citrate.

88 JIANG ET AL.



SFD powder consisted of highly porous particleswith an uneven surface (Fig. 6) and a bulk densityof approximately 0.04 g/cc. The average particlesize was approximately 70 mm, a size appropri-

ate for deposition in the nasal cavity and largeenough to minimize chances of entry into thebronchioles or deep lungs. The lyophilized cakewas converted into a powder by ball milling.Additionally, to study the potential utility of bio-adhesion, selected formulations were preparedcontaining dry chitosan, which was blended withrPA/CpG/trehalose powders prior to filling into INdelivery devices for the animal testing describedpreviously.27

Effect of SFD on rPA Stability

In an attempt to determine any changes in rPAintegrity that might occur due to the SFD process,samples were collected at various stages, includ-ing solution preparation, spraying, instant freez-ing of sprayed droplets, and SFD. The driedpowder was reconstituted for SDS–PAGE analy-sis (Fig. 7), and SEC-HPLC (Fig. 8). Under bothreducing and nonreducing conditions, samplesfrom all four stages appeared as a predominant,intense 83 kDa band identical to the fresh rPAstandard. The CpG oligonucleotide did not haveany noticeable effect. A variety of physical studiesalso found no evidence of interaction between theprotein and oligonucleotide (data not shown). Twoweaker bands of lower molecular weight appearedat approximately 65 kDa and 58 kDa. Comparingthe intensity of the bands, these impurities re-present a concentration of less than 5% of rPAmonomer and exist equally in the rPA standard

Table 3. Effect of Various GRAS Excipients onrPA Aggregation

Excipient

Molar Ratio(Excipient: rPA);

Molarity (Excipient) orwt. Percent (Excipient) % Inhibition*

Brij 35 0.10% 99Sodium citrate 0.2 M 95Sodium citrate 0.1 M 94Trehalose 20.00% 94Malic acid 0.15 M 83Trehalose 10.00% 78Brij 35 0.05% 72Tween 80 0.10% 46Brij 35 0.01% 41Tween 80 0.05% 23Tween 20 0.05% 22Tween 80 0.01% 17Tween 20 0.01% 16Tween 20 0.10% 14Dextran T40 0.1 12Dextran T40 2.5 1Brij 35 0.10% 0Dextran T40 1.0 �3Pluronic F-68 0.10% �12Pluronic F-68 0.05% �14Pluronic F-68 0.01% �19

*Values are�3%.

Figure 4. Protection against rPA aggregation by trehalose at six different concentra-tions: 0%, 2.5%, 5%, 10%, 15%, and 20%.




and all processed samples. No aggregation bandswere observed under nonreducing conditions,suggesting that covalent disulfide linkages werenot formed as a result of the large air–liquid in-terface present during the spaying process. Thus,this analysis suggests that the SFD process yield-ed native-like rPA upon reconstitution.

In SEC chromatograms (Fig. 8), rPA and CpGpeaks were well resolved and accurately quanti-fied. The SFD powder and the liquid prior to theSFD process showed comparable rPA and CpGretention times as well as similar peak areas. Noadditional aggregation or degradation was detect-ed. Therefore, retention of rPA integrity was alsosupported by SEC-HPLC studies.

Storage Stability Studies of Anthrax VaccineFormulations under Accelerated Conditions

Stability testing was conducted at 25 and 408C.The higher temperature was chosen as an accel-erated stability condition, to rapidly obtain datafor extrapolation to longer term storage at lowertemperatures. Stability at ambient conditions ishighly desirable for convenience of use, potentialfor rapid dissemination after a bioterror attack,and to minimize costs for maintenance of a vac-cine stockpile. The SFD powder anthrax vaccineconsists of rPA, CpG, trehalose, and traceamounts of buffer salts. The liquid formulationwas of the same composition and was tested along

Figure 5. Increase in rPA thermal transition temperature in the presence of trehalosecompared to several other excipients as determined by CD (A) and Trp fluorescence (B).

90 JIANG ET AL.



with the powder for comparison and as a control.During incubation at both 25 and 408C, powdersamples maintained their physical appearance asa light powder and could be rapidly reconstitutedinto clear solutions at each time point.

At 258C(Fig. 9A), the stability of thepowder rPAwas not adversely affected throughout the 15 daysof incubation since over 90% intact monomercontent was found at each time point by SEC.The molecular weight (83 kDa) of the monomer

Figure 6. SEM of spray freeze dried rPA anthrax vaccine powder. [Color figure can beseen in the online version of this article, available on the website, www.interscience.wiley.com.]

Figure 7. Retention of rPA integrity throughout the spray freeze drying process asdetermined by SDS–PAGE under reducing conditions (A) and nonreducing conditions(B). Lanes 1 and 8, MW marker; lane 2, CpG Std; lane 3, rPA Std; lane 4, rPA/CpG/trehalose solution; lane 5, solution sprayed from nozzle; lane 6, instant freezing ofsprayed droplets in liquid nitrogen; lane 7, reconstituted spray freeze dried powder.[Color figure can be seen in the online version of this article, available on the website,www.interscience.wiley.com.]




peak was confirmed by light scattering. In con-trast, the liquid formulation displayed significantdegradation over the course of the study. Theresidual stability of the liquid at day 15 was foundto be only 15% (Fig. 9A).

At 408C (Fig. 9B), the powder retained over 90%stability during the entire time course of 29 days.The liquid formulation, however, completely lostits monomer content in 1 day. The intact rPAmonomer peak was below the low level of quanti-fication of the SEC method after only 24 h in-cubation and remained undetectable throughoutthe study. Retention of stability in the powderformulation at 408C suggests that the powdermaybe successfully stored for long periods at roomtemperature. Removal of the need for the coldchain during storage and transportation alongwith the potential for self-administration using INdelivery may provide significant advantages forbiodefense mass immunization.

The SEC results correlate closely with the find-ings of SDS–PAGEunder reducing conditions. Forexample, on day 0, both liquid and powder formu-lations were identical to the fresh rPA standard(Fig. 10A). On day 1, at 408C, the intact rPA

monomer band in the liquid formulationwasmuchlighter than the rPA standard in intensity andmore prominent degradation bands appeared be-tween 30 and 42 kDa (Fig. 10B). It should be notedthat the total intensity of all bands for the solutionsample was still lower than that of the standard,indicating that additional smaller fragments lessthan 4 kDa may have been generated. The liquidformulation continued todegrade further after day1. On day 3, the 83 kDa rPA monomer band wascompletely absent and only a few bands of fadedintensity at 40 kDa and lower were apparent(Fig. 10C). For the SFD powder formulation underthe same conditions, the samples appearunalteredcompared to the rPA standard (Fig. 10A–C). Thispattern persisted through day 29. At 258C, thedegradation of rPA in the liquid formulationappeared to follow the same pattern as that at408C except that the deterioration process wasmore gradual, that is, it took longer to reach thesame degree of degradation at 258C than at 408C.As expected, the powder at 258C retained the samecharacteristics as the intact standard.

To gain a better understanding of the effect oftemperature on the liquid stability of rPA and to

Figure 8. Retention of rPA stability after spray freeze drying as determined by SEC-HPLC.

92 JIANG ET AL.



estimate whether the liquid could be storedrefrigerated instead of frozen, additional liquidstability testing was performed at 208C and 308C.Figure 11A shows loss of intact rPA for the liquidformulation as a function of time at four tempera-tures (208C, 258C, 308C, and 408C) as determinedby SEC-HPLC. As predicted, the rate of deteriora-tion increased as the temperature was elevated,and decreased with time. The stability profile wasfittedusing afirst ordermodel (Fig. 11B)where thelog percentage of intact rPA remainingwas plottedagainst time. This simple model fits reasonablywell for the experimental data at 208C and 258Cwith an R2> 0.96. The 308C profile had only threetime points during the period of rPA degradation,which could account for the poorer fit (R2¼ 0.82).Based on these results, the liquid formulationshould be kept frozen during long-term storage, atleast under these formulation conditions.

SUMMARY AND CONCLUSIONS

We previously reported that IN anthrax vaccinepowders provided complete protection of rabbitsagainst anthrax lethal aerosol challenge. Theformulation development for this vaccine powderis comprehensively described in this study. Thiswork first investigated the physical stability of

Figure 9. Storage stability of rPA liquid and sprayfreeze dried powder at 258C (A) and 408C (B) determinedby SEC-HPLC.

Figure 10. Degradation of the liquid formulation at 408C determined by SDS–PAGEunder reducing conditions on day 0 (A; lane 1, 2 mg rPA Std; lane 2, MWmarker; lane 3,2 mg CpG Std; lane 4 and 5, liquid formulation containing rPA 2 mg and 0.5 mg,respectively; lane 6 and 7, SFD powder formulation containing rPA 2 mg and 0.5 mg,respectively), day 1 and day 3 (B andC; lane 1,MWmarker; lane 2, rPAStd 1.5 mg; lane 3,CpG Std 1.5 mg; lane 4, liquid formulation containing rPA 1.5 mg; lane 5, SFD powderformulation containing rPA 1.5 mg). [Color figure can be seen in the online version of thisarticle, available on the website, www.interscience.wiley.com.]




rPA in solution as a function of pH and tempera-ture. The stabilizing effect of GRAS excipentson rPA aggregation and loss of conformationalstability was determined. Based on these find-ings, anthrax vaccine powders including SFD orFD dosage forms were prepared at pH 7–8 withtrehalose as a stabilizer and bulking agent. Thestability of rPA was retained after powder pro-cessing. The dry powder also maintained proteinintegrity under both ambient and acceleratedaging temperatures for approximately 1 monthwhile the liquid formulation showed rapid degra-dation. Such degradation appears to follow firstorder kinetics.

Based on these studies, stable IN powderformulations of rPA were developed as a potentialalternative to heat-labile liquid formulationsand conventional parenteral delivery by injec-tion. Combination of this powder formulationwith the noninvasive delivery platform describedpreviously27,31 could greatly improve the viabi-lity of mass immunization for this importantpathogen.

ACKNOWLEDGMENTS

The authors thank Harry Sugg for performingelectron microscopy on powder samples, and PatMcCutchen and Linda Tingen for administrativeassistance in preparing the manuscript. Financialsupport was provided by funding from the U.S.Army Medical Research and Material Command,agreement number DAMD17-03-2-0037.

REFERENCES

1. Oncu S, Sakarya S. 2003. Anthrax—An overview.Med Sci Monit 9:RA276–RA283.

2. Flick-Smith HC, Eyles JE, Hebdon R, Waters EL,Beedham RJ, Stagg TJ, Miller J, Alpar HO, BaillieLWJ, Williamson ED. 2002. Mucosal or paren-teral administration of microsphere-associatedBacillus anthracis protective antigen protectsagainst anthrax infection in mice. Infect Immun70:2022–2028.

3. Gupta P, Singh S, Tiwari A, Bhat R, Bhatnagar R.2001. Effect of pH on stability of anthrax lethalfactor: Correlation between denaturation and activ-ity. Biochem Biophys Res Commun 284:568–573.

4. Bhatnagar R, Batra S. 2001. Anthrax toxin. CritRev Microbiol 27:167–200.

5. Quinn CP, Turnbull PC. 1998. Anthrax. In: CollierL, Barlow A, Sussman M, editors. Topley andWilson’s microbiology and microbial infections,9th edition. London, UK: Arnold.

6. Tan Y, Hackett NR, Boyer JL, Crystal RG. 2003.Protective immunity evoked against anthrax lethaltoxin after a single intramuscular administrationof an adenovirus-based vaccine encoding huma-nized protective antigen. HumGene Ther 14:1673–1682.

7. Abrami L, Lindsay M, Parton RG, Leppla SH,van der Goot FG. 2004. Membrane insertionof anthrax protective antigen and cytoplasmicdelivery of lethal factor occur at different stagesof the endocytic pathway. J Cell Biol 166:645–651.

8. Arora N, Klimpel KR, Singh Y, Leppla SH. 1992.Fusions of anthrax toxin lethal factor to the ADP-ribosylation domain of Pseudomonas exotoxin Aare potent cytotoxins which are translocated to thecytosol of mammalian cells. J Biol Chem 267:15542–15548.

9. Chauhan V, Bhatnagar R. 2002. Identification ofamino acid residues of anthrax protective antigeninvolved in binding with lethal factor. InfectImmun 70:4477–4484.

10. Pezard C, Berche P, Mock M. 1991. Contribu-tion of individual toxin components to virulenceof Bacillus anthracis. Infect Immun 59:3472–3477.

Figure 11. Degradation profiles of the liquid formu-lation at 20, 25, 30, and 408C as determined by SEC-HPLC (A) and the data fit to 1st order kinetics (B).

94 JIANG ET AL.



11. Hanna P. 1999. Lethal toxin actions and theirconsequences. J Appl Microbiol 87:285–287.

12. Leppla SH. 1982. Anthrax toxin edema factor: Abacterial adenylate cyclase that increases cyclicAMP concentrations of eukaryotic cells. Proc NatlAcad Sci USA 79:3162–3166.

13. Hoover DL, Friedlander AM, Rogers LC, Yoon IK,Warren RL, Cross AS. 1994. Anthrax edema toxindifferentially regulates lipopolysaccharide-inducedmonocyte production of tumor necrosis factor alphaand interleukin-6 by increasing intracellular cyclicAMP. Infect Immun 62:4432–4439.

14. Collier RJ, Young JA. 2003. Anthrax toxin. AnnuRev Cell Dev Biol 19:45–70.

15. Mogridge J, Cunningham K, Collier RJ. 2002. Sto-ichiometry of anthrax toxin complexes. Biochem-istry 41:1079–1082.

16. Prokupek K, Dvorak R, Polacek R. 1981. Immuno-genic effect of anthrax protective antigen. Vet Med(Praha) 26:279–290.

17. Ivins BE, Fellows PF, Nelson GO. 1994. Efficacy ofa standard human anthrax vaccine against Bacil-lus anthracis spore challenge in guinea-pigs.Vaccine 12:872–874.

18. Jendrek S, Little SF, Hem S, Mitra G, GiardinaS. 2003. Evaluation of the compatibility of asecond generation recombinant anthrax vaccinewith aluminum-containing adjuvants. Vaccine 21:3011–3018.

19. Baillie L, Hebdon R, Flick-Smith H, Williamson D.2003. Characterisation of the immune response tothe UK human anthrax vaccine. FEMS ImmunolMed Microbiol 36:83–86.

20. Reuveny S, White MD, Adar YY, Kafri Y, AltboumZ, Gozes Y, Kobiler D, Shafferman A, Velan B.2001. Search for correlates of protective immunityconferred by anthrax vaccine. Infect Immun 69:2888–2893.

21. Ivins BE, Welkos SL. 1986. Cloning and expres-sion of the Bacillus anthracis protective antigengene in Bacillus subtilis. Infect Immun 54:537–542.

22. Singh Y, Ivins BE, Leppla SH. 1998. Study ofimmunization against anthrax with the purifiedrecombinant protective antigen of Bacillus anthra-cis. Infect Immun 66:3447–3448.

23. Clinicaltrials.gov website www.clinicaltrials.gov/ct/show/NCT00057525. ed.

24. Tebbey PW, Scheuer CA, Peek JA, Zhu D, LaPierreNA, Green BA, Phillips ED, Ibraghimov AR,Eldridge JH, Hancock GE. 2000. Effective mucosalimmunization against respiratory syncytial virususing purified F protein and a genetically detoxifiedcholera holotoxin, CT-E29H. Vaccine 18:2723–2734.

25. Papp Z, Babiuk LA, Baca-Estrada ME. 1999.Antigen-specific cytokine and antibody isotypeprofiles induced by mucosal and systemic immu-

nization with recombinant adenoviruses. ViralImmunol 12:107–116.

26. IsakaM, Yasuda Y, Taniguchi T, Kozuka S,MatanoK, Maeyama J, Morokuma K, Ohkuma K, Goto N,Tochikubo K. 2003. Mucosal and systemic antibodyresponses against an acellular pertussis vaccine inmice after intranasal co-administration with re-combinant cholera toxin B subunit as an adjuvant.Vaccine 21:1165–1173.

27. Mikszta JA, Sullivan VJ, Dean C, Waterston AM,Alarcon JB, Dekker JP, Brittingham JM, Huang J,Hwang CR, Ferriter M, Jiang G, Mar K, SaikhKU, Stiles BJ, Roy CJ, Ulrich RG, Harvey NG.2005. Protective immunization against inhala-tional anthrax: A comparison of minimally-invasivedelivery platforms. J Infect Dis 191:278–288.

28. Kueltzo LA, Ersoy B, Ralston JP, Middaugh CR.2003. Derivative absorbance spectroscopy andprotein phase diagrams as tools for comprehensiveprotein characterization: A bGCSF case study.J Pharm Sci 92:1805–1820.

29. Klinman DM. 2003. CpG DNA as a vaccineadjuvant. Expert Review of Vaccines 2:305–315.

30. Kueltzo LA, Middaugh CR. 2003. Structural char-acterization of bovine granulocyte colony stimulat-ing factor: Effect of temperature and pH. J PharmSci 92:1793–1804.

31. Huang J, Garmise RJ, Crowder TM, Mar KM,Hwang CR, Hickey AJ, Mikszta JM. 2004. SVJ2004. A novel dry powder influenza vaccine andintranasal delivery technology: Induction of sys-temic and mucosal immune responses in rats.Vaccine 23:794–801.

32. Radha C, Salotra P, Bhat R, Bhatnagar R. 1996.Thermostabilization of protective antigen—Thebinding component of anthrax lethal toxin. JBiotechnol 50:235–242.

33. Buchner J, RennerM, Lilie H, Hinz HJ, Jaenicke R,Kiefhabel T, Rudolph R. 1991. Alternatively foldedstates of an immunoglobulin. Biochemistry 30:6922–6929.

34. Mayo KH, Barker S, Kuranda MJ, Hunt AJ, MyersJA, Maione TE. 1992. Molten globule monomerto condensed dimer: Role of disulfide bonds inplatelet factor-4 folding and subunit association.Biochemistry 31:12255–12265.

35. Tanksale A, Ghatge M, Deshpande V. 2002. Alpha-crystallin binds to the aggregation-prone molten-globule state of alkaline protease: Implicationsfor preventing irreversible thermal denaturation.Protein Sci 11:1720–1728.

36. Karlsson M, Martensson LG, Olofsson P, CarlssonU. 2004. Circumnavigating misfolding traps in theenergy landscape through protein engineering:Suppression of molten globule and aggregation incarbonic anhydrase. Biochemistry 43:6803–6807.

37. Asghari SM, Khajeh K, Ranjbar B, Sajedi RH,Naderi-Manesh H. 2004. Comparative studies on




trifluoroethanol (TFE) state of a thermophilicalpha-amylase and itsmesophilic counterpart: Lim-ited proteolysis, conformational analysis, aggrega-tion and reactivation of the enzymes. Int J BiolMacromol 34:173–179.

38. Mazon H, Marcillat O, Forest E, Smith DL, Vial C.2004. Conformational dynamics of the GdmHCl-induced molten globule state of creatine kinase

monitored by hydrogen exchange andmass spectro-metry. Biochemistry 43:5045–5054.

39. Almstedt K, Lundqvist M, Carlsson J, Karlsson M,Persson B, Jonsson BH, Carlsson U, HammarstromP. 2004. Unfolding a folding disease: Folding,misfolding and aggregation of the marble brainsyndrome-associated mutant H107Y of humancarbonic anhydrase II. J Mol Biol 342:619–633.

96 JIANG ET AL.



Documents

Anthrax vaccine powder formulations for nasal mucosal delivery · 5a. CONTRACT NUMBER 5b. GRANT NUMBER 5c. PROGRAM ELEMENT NUMBER 6. AUTHOR(S) Jiang /G Joshi, SB Peek, LJ Brandau,